[NIFI-6417] JOLT processors should accept more character encodings

• Type: Improvement
• Status: Patch Available
• Priority: Major
• Resolution: Unresolved
• Component/s: Extensions
• Assignee: Travis Neeley
• Reporter: Matt Burgess
• Created: 02/Jul/19 20:42
• Updated: 11/May/21 00:07

Description

Currently JoltTransformJSON and JoltTransformRecord are hard-coded to input and output UTF-8 encoding. These processors should be able to at least accept a user-configured input encoding such as UTF-16, ISO-8859-1, etc.

We could retain the behavior of writing out in UTF-8, but I think for consistency and flexibility we should add an Output Character Encoding property as well. Both would be defaulted to UTF-8 to maintain current default behavior.
NOTE: There may be a limitation at present with JoltTransformRecord regarding String fields, where they are hard-coded to be UTF-8. If that is the case, then both processors should still offer the Input Character Encoding, but only JoltTransformJSON would have the Output Character Encoding property.

Issue Links

• links to: GitHub Pull Request #3576